Thrombolytic therapy of acute ischemic stroke during pregnancy

Increased risk of stroke during pregnancy is controversial,^1,2 and there is inadequate information regarding the use of thrombolytics in pregnant women. The package insert for recombinant human tissue plasminogen activator (rt-PA) (Activase) denotes pregnancy to be a relative contraindication for administration of the drug and labels rt-PA as pregnancy category C (uncertain safety). Others denote it to be pregnancy category B (presumed safety based on animal studies).³ Urokinase (UK) is classified as pregnancy category B. We report a series of eight patients who underwent thrombolysis for acute ischemic stroke (AIS) during pregnancy. We also attempt to clarify the maternal safety and efficacy of rt-PA in pregnant women.

Methods.

When a pregnant woman presented within 3 hours of the onset of AIS, we communicated with 96 colleagues at 40 medical centers in four countries regarding whether any had treated a pregnant woman with rt-PA for AIS. Colleagues included those with a professed academic and clinical interest in AIS therapy, or neurologic and neurosurgical emergency and critical care. We received notice of eight other cases from this selected group of physicians. One of these cases was unavailable for report here.

We queried contributors regarding clinical presentation, NIH Stroke Scale (NIHSS) score, medical history, suspected etiology of the index stroke, details of treatment course (including complications, if any), and clinical outcomes of mothers and fetuses. The NIHSS was recorded before and after treatment and at the most recent available patient follow-up.

We searched the National Library of Medicine's MEDLINE database from 1966 until June 2005 using the medical subject headings of thrombolytics in pregnancy, rt-PA in pregnancy, and stroke in pregnancy. Additionally, we queried Genentech for any reports to the company regarding the use of this drug for ALS in pregnant women.

Results.

We identified a total of eight women treated with thrombolytics for AIS during pregnancy. Specific details are summarized in the table. Brief vignettes of the eight patients are available (appendix E-1, available on the Neurology Web site at www.neurology.org). Genentech, Inc. denoted no reports to the manufacturer of rt-PA use for AIS during pregnancy.

The average maternal age was 32 years, and the mean gestational age was 11 weeks, ranging from 4 to 37 weeks. Seven of the mothers were in their first trimester, and one was in the third trimester. The etiologies of their strokes were discontinuation of anticoagulants in two mechanical heart valve patients because of nausea in one and inability to afford low-molecular-weight heparin in the other; and hypercoagulable state in five patients, with a patent foramen ovale in one of these women and polycythemia vera with thrombocytosis in another. One patient had bacterial endocarditis. Three patients received IV rt-PA; in two of these, the treating physicians were unaware of the pregnancy, and of these two, one received rt-PA because of time constraints before the results of a urine pregnancy test were available from the laboratory, and in the other patient, a pregnancy test was not performed because of the patient's miscalculation of her last menstrual period. One patient received intra-arterial rt-PA in an effort to reduce fetal exposure. Four patients received UK, one for midbasilar total occlusion, one locally into the left middle cerebral artery, and two into the cerebral venous sinuses for cerebral venous thrombosis.

With one exception, mothers recovered well from their strokes. Adverse outcomes in these eight mothers included one intrauterine hematoma (drained), one buttock hematoma (managed conservatively), two asymptomatic small intracranial hemorrhages (one at a catheter site and one intraparenchymal, both after local UK), and one fatal malignant infarction as a result of arterial dissection complicating angioplasty (representing the only poor maternal outcome). This death cannot be attributed solely and directly to thrombolytic therapy. There were no symptomatic intracerebral hemorrhages (ICHs). Of the seven surviving mothers (87.5%), three elected to have therapeutic abortions. There were two miscarriages, both during the first trimester, 2 to 3 days after UK administration. Autopsy in one of these fetuses revealed lethal chromosomal anomalies. Autopsy was not performed in the other fetus, and it was presumed to be a spontaneous first trimester abortion. Two babies were delivered at term without any complications (one each intra-arterial [IA] rt-PA and IA UK).

The literature review identified 175 pregnant women treated with any thrombolytic, all but 3 for diseases other than stroke. We identified 12 pregnant women, with enough useful details reported, who were treated specifically with rt-PA for myocardial infarction and thromboembolic diseases other than stroke. Of the 12 fetuses, 6 (50%) were healthy after delivery, 1 died after delivery (see below), and 1 died because of emergent medical termination of pregnancy; in 3 of the reports, fetal outcome was not clearly stated. Overall, 6 of the 12 babies were delivered preterm electively based on maternal interest, and all of these babies, except 1, were healthy. One preterm baby delivered at 35 weeks of gestation, by caesarean delivery, died on day 14 of respiratory distress syndrome; autopsy revealed cerebral, cerebellar, and subarachnoid hemorrhages (12 full references available in appendix E-2).

Of the three cases reporting intra-arterial rt-PA treatment for stroke during pregnancy,^4–6 two were first trimester,^4,5 and one was third trimester.⁶ Two women had marked neurologic improvement,^4,6 and one initially improved (recanalized the M1 segment), then worsened (small ICH), then further deteriorated (larger ICH), and then partially improved.⁵ All had normal babies delivered.

Discussion.

Based on these case summaries and reports obtained from literature, it seems that pregnant women generally can be safely treated with IV or IA rt-PA, or UK for AIS and can (but not always) have reasonably good outcomes. Effects on fetal viability and any potential complications of thrombolytics (e.g., teratogenicity) remain largely unknown. This series of eight patients is too small a number to draw any major efficacy or safety conclusions.

The major obstetric concern regarding the use of thrombolytics during pregnancy is their effect on the placenta, possibly resulting in premature labor, placental abruption, or fetal demise. A summary of the literature regarding thrombolytic (rt-PA, urokinase, etc.) use in pregnancy identified 172 pregnant women affected with thromboembolic conditions who were treated with thrombolytics⁷; however, none of these patients were treated for stroke. In that review, 5% to 8% of deliveries from patients treated with thrombolytics occurred preterm, which is within the reported 10% incidence of all preterm deliveries in the United States.⁸

Other concerns include hemorrhage during parturition or caesarean delivery. This risk exists if the patient goes into labor within the context sensitive half-time (the time of the drug's effective physiologic and pharmacologic effects in the setting of the disease process being studied) after administration of the thrombolytic.

Regarding teratogenicity, there are no data in the literature or any reports to the manufacturer regarding this issue, either for rt-PA or for UK. One might consider that the teratogenic effect, if any, might be less with IA rt-PA use, because the dose of IA rt-PA is less than the IV dose.

The ethical situation of a young woman being less disabled as a result of thrombolytics for AIS vs the risk (teratogenicity, prematurity, death) of the unborn baby must be considered in each clinical situation.

Acknowledgment

The authors thank J.J. King, K. Smoot, and M.E. Redman for assistance with cases. They also thank Dr. Jose Biller for supplying references pertinent to the investigation.